Hi, my guess is that ASN decoding isn't a major performance factor. We do so much stuff with the messages once they are decoded (including database backends etc.) that this shouldn't be the limiting factor. Also it would be huge effort, basically replacing OpenH323. If we were to optimize GnuGk for performance it would be nice to have a testbed first, that generates messages for both call ends. That way we could have repeatable profiler diagnostics and measure the impact of possible optimizations. Measuring should really be the first step. Its easy to be wrong where the time is spent. (And of course that includes my guess that ASN is not a factor.) Regards, Jan cn_zhx wrote: > Hi all, > According to gnugk's statement, it's call processing capability is > 30 CAPS which mean 30 new calls can be connected per second, this value is > much lower than SIP system. According to OpenSER's statement, their CAPS is > more than one hundred. Why ? > I guess maybe openh323's ASN parser is too slow. Which affect the > performance of gnugk. So I suggest us this ASN parser > (http://iiiasn1.sourceforge.net). -- Jan Willamowius, jan@xxxxxxxxxxxxxx, http://www.gnugk.org/ ------------------------------------------------------------------------- Using Tomcat but need to do more? Need to support web services, security? Get stuff done quickly with pre-integrated technology to make your job easier. Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642 _______________________________________________________ Posting: mailto:Openh323gk-users@xxxxxxxxxxxxxxxxxxxxx Archive: http://sourceforge.net/mailarchive/forum.php?forum_id=8549 Unsubscribe: http://lists.sourceforge.net/lists/listinfo/openh323gk-users Homepage: http://www.gnugk.org/
